1 Projecting marine fish production and catch potential in Bangladesh in the 21 st century 2 under long-term environmental change and management scenarios
3 Jose A. Fernandes 1, Susan Kay 1, Mostafa A.R. Hossain 2, Munir Ahmed 3, William W.L. 4 Cheung 4, Attila N. Lazar 5, Manuel Barange 1
5
6
7 1. Plymouth Marine Laboratory, Prospect Place, The Hoe, Plymouth, U.K. PL13 DH
8 2. Department of Fish, Biology and Genetics. Bangladesh Agricultural University, 9 Mymensingh 2202, Bangladesh.
10 3. TARA, 1 Purbachal Road, Nartheast Badda, Dhaka 1212, Bangladesh
11 4. Fisheries Centre, AERL 2202 Main Mall, The University of British Columbia, Vancouver, 12 B.C., Canada, V6T 1Z4
13 5. Faculty of Engineering and Environment, University of Southampton, University Road, 14 Southampton, Hampshire, United Kingdom, SO17 1BJ.
15
16 17 18 Abstract
19
20 The fisheries sector is crucial to the Bangladeshi economy and wellbeing, accounting for
21 4.4% of national Gross Domestic Product (GDP) and 22.8% of agriculture sector production,
22 and supplying ca. 60% of the national animal protein intake. Fish is vital to the 16 million
23 Bangladeshis living near the coast, a number that has doubled since the 1980s. Here we
24 develop and apply tools to project the long term productive capacity of Bangladesh marine
25 fisheries under climate and fisheries management scenarios, based on downscaling a global
26 climate model, using associated river flow and nutrient loading estimates, projecting high
27 resolution changes in physical and biochemical ocean properties, and eventually projecting
28 fish production and catch potential under different fishing mortality targets. We place
29 particular interest on Hilsa shad (Tenualosa ilisha ), which accounts for ca. 11% of total
30 catches, and Bombay duck (Harpadon nehereus ), a low price fish that is the second highest
31 catch in Bangladesh and is highly consumed by low income communities. It is concluded that
32 the impacts of climate change, under greenhouse emissions scenario A1B, are likely to reduce
33 the potential fish production in the Bangladesh Exclusive Economic Zone (EEZ) by less than
34 10%. However, these impacts are larger for the two target species. Under sustainable
35 management practices we expect Hilsa shad catches to show a minor decline in potential
36 catch by 2030 but a significant (25%) decline by 2060. However, if overexploitation is
37 allowed catches are projected to fall much further, by almost 95% by 2060, compared to the
38 Business as Usual scenario for the start of the 21 st century. For Bombay duck, potential
39 catches by 2060 under sustainable scenarios will produce a decline of less than 20%
40 compared to current catches. The results demonstrate that management can mitigate or
41 exacerbate the effects of climate change on ecosystem productivity. 42
43 Introduction
44 Bangladesh is recognized as being highly vulnerable to the potential impacts of global and
45 regional climate change due to its geographic location and low-lying topography, putting it at
46 risk of extreme flooding and weather events. The Bangladesh delta region is home to over
47 500 million of some of the poorest people worldwide (FAO, 2007), who rely on the rich
48 ecosystem services available (Newton et al., 2007; Shamsuddoha and Chowdhury, 2007). The
49 coastal population of Bangladesh has doubled since the 1980s, now reaching in excess of 16
50 million (~ 10 % of the total population), most of them experiencing both poverty and
51 environmental vulnerability. It is known that in economic systems close to or below the
52 poverty line, both subsistence and cash elements of the economy rely disproportionately on
53 ecosystem services (Newton et al., 2007; Allison, 2009). Understanding the effects of climate
54 change on ecosystem services is thus particularly relevant in poor regions (Nicholls et al.,
55 2015).
56
57 Fisheries play an important provisioning service in Bangladesh, supplying 60% of the protein
58 intake (DoF, 2013). Fisheries catches can be divided into marine ( ca. 17%), inland open
59 water or captured ( ca. 28%) and closed water or cultured ( ca. 55%). Inland open water
60 includes river and estuaries, the Sundarbans mangrove area, shallow marshy lakes known as
61 Beels, Kaptai Lake and seasonally flooded plains. Inland closed water corresponds to ponds,
62 semi-enclosed water bodies, lakes and shrimp/prawn farms. Although cultured fisheries
63 dominate the catch data, these depend strongly on river and marine ecosystems for the
64 provision of larvae and juveniles. Marine catches come from the Bangladesh Exclusive
65 Economic Zone (EEZ), which covers an area of 86,392 km 2, and represent an important
66 economic activity, with 0.55 Mt of fish production in 2011. About 225 trawlers and 52,514 67 mechanized and non-mechanized boats are engaged in fishing (DoF, 2013). Climate change
68 has been predicted to decrease the productive potential of fisheries in South and Southeast
69 Asia (Barange et al., 2014). Understanding how this impact translates into the future
70 provision of fish products in Bangladesh is crucial for the sustainability of fisheries
71 dependent communities in coming decades.
72
73 In this paper we assess the impacts of climate change on Bangladesh’s fish resources by
74 quantifying the cumulative physical, biological and ecological impacts in the EEZ of
75 Bangladesh over the 21 st century, and how these affect the fish production potential in the
76 coastal zone. We include a set of contrasting management and exploitation scenarios for the
77 EEZ of Bangladesh to project plausible scenarios of fish production by 2060 that combine
78 sustainable management and environmental impacts. These scenarios are specifically focused
79 on the two species that provide the largest marine and inland catches, Hilsa shad, Tenualosa
80 ilisha, and Bombay duck, Harpadon nehereus. The management scenarios are intended to
81 inform fisheries managers in Bangladesh in the pursuit of sustainable management strategies
82 under a future dominated by climate change.
83
84 Hilsa shad is a euryhaline anadromous shad found in marine, coastal and freshwater
85 environments, often schooling in coastal waters. Hilsa is the single most important fish
86 species in Bangladesh, accounting for more than 10% of the total national fish production,
87 and responsible for about 1% of Bangladesh’s GDP. About 460,000 fishers of 148 Upazilas
88 (sub-districts) are directly employed in Hilsa fishing with an indirect employment of about
89 2.5 million people in the wider Hilsa sector (trading, processing etc.). Bombay duck is the
90 second highest catch in Bangladesh and a much cheaper fish than Hilsa Shad ( approx. 65
91 BDT.kg -1 vs 430 BDT.kg -1, unpublished data). As it is cheap and still caught in abundance, 92 Bombay duck, currently, is one of the most preferred fish for poor and middle class
93 consumers all over the coast (on average, 14% of all daily sales is Bombay duck). Dried
94 Bombay duck is also regarded as the number one dried fish and is very popular, particularly
95 in the eastern part of the coast. It is a very soft and highly perishable table fish and also
96 valuable in laminated or dried from.
97
98 Growth of human populations and changes in global consumption patterns will continue to
99 place heavy demands on fish populations (Delgado et al., 2003), particularly in fish-
100 dependent regions like Bangladesh (Toufique and Belton 2014). Here we used quantitative
101 scenarios of climate and socio-economic changes in the Bangladeshi marine ecosystem and
102 its major fish species to assist in the development of adaptation measures.
103
104 105 106 Material and Methods
107
108 In order to provide estimates of fish production potential under climate change we used a
109 combination of atmospheric, hydrological, ocean circulation and ocean biogeochemical
110 models, driving changes in ocean productivity and fisheries potential (Fig. 1). These models
111 produce yearly time steps and spatially resolved results. The Bay of Bengal (BoB) physico-
112 biogeochemical model simulates the cycling of the main nutrients through the benthic and
113 planktonic pelagic ecosystems. Outputs from this model drive two fisheries models: a size-
114 spectrum model to provide time-series of total marine fish production by size, and a species-
115 based model to compute total marine production of the main species by size. The effects of
116 human activities and fisheries management policies were explored through fishing mortality
117 scenarios. Details of each model are provided below.
118
119 Climate and hydrological models
120 Previous climate modelling studies in this region have tended to focus on the wider Indian
121 subcontinent rather than Bangladesh. Most studies project a generic increase in atmospheric
122 temperature, annual rainfall and heavy precipitation events. Climate data for this study was
123 taken from the UK Met Office regional climate model (RCM) HadRM3P, which is
124 dynamically downscaled from the global circulation model HadCM3 (Caesar et al., 2015).
125 The greenhouse gas emission scenario used here is the Special Report on Emissions
126 Scenarios (SRES) A1B (IPCC 2007), a medium-high emissions scenario developed for
127 Intergovernmental Panel on Climate Change (IPCC) 3rd and 4 th Assessment Reports, which
128 still underpins research into climate impacts. In order to capture some of the model
129 uncertainties we considered three different climate projections from a 17-member ensemble
130 of HadCM3 runs (Caesar et al 2015). The three climate model runs selected correspond to a 131 range of possible future outcomes for the Bay of Bengal, between the standard (Q0), drier
132 and warmer (Q16) and intermediate rainfall and temperature (Q8) projections (Caesar et al.,
133 2015). Q0, Q8 and Q16 have successively higher sensitivity to greenhouse gas forcing
134 because of the different parameter values used in the general circulation model for these
135 ensemble members. As delta regions are particularly sensitive to precipitation and river run-
136 off, outputs from an Integrated Catchment Model (INCA,Whitehead et al., 2015a,b) were
137 used to determine run-off and associated nutrient loadings from the delta rivers into the BoB
138 for each projection. The model simulates factors controlling flow and water quality dynamics
139 in both land and stream components of river catchments. The INCA model application took
140 account of both climatic scenarios (Q0, Q8, Q16) and patterns of upstream water use
141 according to three socioeconomic scenarios (Less Sustainable, LS, Business as Usual, BaU,
142 More Sustainable, MS) scenarios (Whitehead et al., 2015a,b). We used the results of the Q0-
143 BAU, Q8-LS and Q16-MS INCA simulation runs to capture the variation of the simulated
144 river flows and nutrient loads (Table I). The rivers in the Ganges-Brahmaputra-Meghna delta
145 region (Fig. 2) account for 40% of flow into the model domain. For all other rivers, for which
146 INCA data was not available, data was extracted from global databases (Global NEWS ,
147 http://marine.rutgers.edu/globalnews/datasets.htm and Dai and Trenberth Global River Flow
148 and Continental Discharge Dataset, http://www.cgd.ucar.edu/cas/catalog/surface/dai-
149 runoff/index.html ).
150
151 Ocean dynamics and biogeochemistry
152 A regional POLCOMS-ERSEM coupled model (Holt and James, 2001; Blackford et al. 2004;
153 Holt et al., 2009) was used to project both the physical state of the ocean (temperature,
154 salinity, currents, light level), and the biogeochemistry and lower trophic levels of the marine
155 food-web in the BoB. The model simulates four phytoplankton functional types, three 156 zooplankton functional types and bacteria, as well as three size classes of particulate organic
157 matter and dissolved and semi-labile organic matter. Four nutrients (C, P, N and Si) are
158 explicitly tracked within the model. The model domain covers the coastal area of the whole
159 BoB (77° to 104° W, 1.3° S to 23° N), and its width is from the coast to 200 km beyond the
160 edge of the continental shelf (Fig. 2). The model uses a rectangular grid with a horizontal
161 resolution of 0.1° and 42 vertical levels distributed according to bottom depth. At the
162 atmospheric boundary it was forced using 3-hourly and daily outputs from the HadRM3P
163 regional climate model described above, and physical conditions at the open ocean boundary
164 were set using monthly outputs from the HadCM3 GCM.
165
166 Nutrient values at the ocean boundary were fixed to values from the World Ocean Atlas
167 (Garcia et al., 2010); since future projections are not available for these variables the values
168 were kept constant during the run. The boundary is advective, so although nutrient values are
169 kept constant nutrient losses and gains at the boundaries are allowed. Keeping nutrient levels
170 at the boundary fixed could have some effect on primary production, but since production
171 here is very low (Martin and Shaji, 2015) the effect is likely to be small compared to changes
172 in more productive zones nearer the coast.
173
174 For each climate dataset, the model was run continuously for 1971-2099. Model outputs
175 including temperature, salinity, current speeds, primary production, dissolved oxygen, pH and
176 plankton biomass were recorded at daily intervals and used to run the fish production models.
177
178 The BoB POLCOMS-EREM model was validated by comparing model outputs to in situ
179 measurements of temperature and salinity and to satellite values of surface chlorophyll. 180 Temperature and salinity data were taken from the World Ocean Database (Boyer et al. 2013),
181 using all available data for the model domain for the period 1993 to 2009. Monthly-
182 aggregated satellite chlorophyll data for 1997-2009 were taken from the database of the
183 Ocean Colour Climate Change Initiative (Hollmann et al., 2013; http://www.esa-oceancolour-
184 cci.org). For validation purposes the model was run with forcing from reanalysis data
185 (ERAinterim, Dee et al., 2011, and GLORYS, Ferry et al., 2012).
186
187
188 Fish production models
189 Outputs from the POLCOMS-ERSEM model were used to drive a dynamic marine
190 ecosystem model that explicitly accounts for food web interactions by linking primary
191 production to fish production through predation. The model estimates potential for fish
192 production by size class, taking in to account temperature effects on the feeding and intrinsic
193 mortality rates of organisms (Blanchard et al. 2012), and hence can make climate-driven
194 projections of changes in potential fish production. Size-based methods like this capture the
195 properties of food webs that describe energy flux and production at a particular size,
196 independent of species’ ecology (Barange et al. 2014).
197
198 To make projections for key species, we used a Dynamic Bioclimate Envelope Model
199 (DBEM), a combined mechanistic-statistical approach that has been applied to a large
200 number of marine species globally (Cheung et al., 2009). The DBEM projects changes in
201 species distribution and abundance while explicitly considering known mechanisms of
202 population dynamics and dispersal (both larval and adult), as well as ecophysiological
203 changes caused by changing ocean conditions (Cheung et al. 2011). Specifically, we 204 employed the SS-DBEM version of the DBEM that incorporates species interactions based
205 on size-spectrum theory and habitat suitability (Fernandes et al., 2013). In SS-DBEM, current
206 distributions of the studied species (Hilsa shad and Bombay duck) are first estimated using
207 the Sea Around Us database method (Close et al. 2006). Then the suitability of each species
208 to different environmental conditions was defined, using its model inferred environmental
209 preference profile (Cheung et al., 2008). Combining ocean dynamics with mortality, growth
210 and dispersal processes the model projects future patterns in distribution, biomass and
211 potential catch (see Cheung et al 2008, 2009 for more details). We applied the size-spectrum
212 model to explore potential changes in the total productivity of the Bangladesh Exclusive
213 Economic Zone under both climate change and fishing scenarios and the SS-DBEM for the
214 two target species (Hilsa shad and Bombay duck).
215
216 Fisheries Management scenarios
217 The fisheries scenarios considered are based on the ecosystem carrying capacity of the
218 Bangladesh Exclusive Economic Zone (EEZ) and aim to provide trends of fish potential for
219 particular species by size class according to specific levels of fishing pressure in relation to
220 the species’ maximum sustainable yield (MSY). MSY is defined as the highest average
221 theoretical equilibrium catch that can be continuously taken from a stock under average
222 environmental conditions (Hilborn & Walters 1992). Based on a simple logistic population
223 growth function and under equilibrium conditions, MSY can be defined as:
224 MSY = B ∞ * intR /4
225 where intR is the intrinsic rate of population increase and B ∞ is the biomass at carrying
226 capacity (Schaefer 1954, Sparre and Venema 1992). In our application, the intR values are 227 calculated based on natural mortality (Pauly 1980; Cheung et al., 2008). This is an
228 approximation and not as reliable as estimates of biomass using survey-based methods
229 (McAllister et al. 2001; Pauly et al., 2013). However, these estimates have proven to be
230 significantly correlated with those from aggregated stock assessments (Froese et al., 2012;
231 Fernandes et al., 2013).
232 Fishing mortality (F m) scenarios were defined by comparing F m estimates for Hilsa shad from
233 the literature with the modelled fishing mortality associated with MSY. Our modelled
234 averaged F MSY is ca. 0.6, which contrasts with much higher levels of fishing mortality in the
235 literature (Haldar and Amin 2005, Amin et al. 2008, Ahmed et al., 2008, Rahman et al.,
236 2012), which would be consistent with expectations of significant overfishing in the region
237 (Khan, 2007; Hussain and Hoq, 2010; Dutta et al., 2012). Thus, we considered three fishing
238 mortality scenarios to provide fish catch and biomass projections as follows:
239 a) Sustainability scenario (MSY): Fishing effort consistent with average F MSY (0.6). This
240 is the theoretical value that results in maximum catches while maintaining the
241 population at their productivity peak.
242 b) Business as Usual scenario (BaU): Fishing mortality consistent with the average of
243 recent estimates of F m (FBaU ). The average fishing mortality for Hilsa shad, the largest
244 fishery in the country, is 1.86, or 3*F MSY . We will consider that Hilsa shad and many
245 of the brackish species in the BoB are being exploited at that rate.
246 c) Overfishing scenario (F OF): Corresponds to a scenario where management is not a
247 constraint to the fishery. Initial runs indicate collapse of catches and biomass at 4 *
248 MSY (F m of 2.4). This is the highest exploitation scenario that we can consider.
249 250
251 Results
252 Correlation between modelled and observed values is high for temperature and salinity (0.97
253 and 0.87 respectively). Chlorophyll was observed to be consistently over-estimated compared
254 to satellite values in the region (validation not shown), but the spatial and seasonal patterns of
255 variation reproduced well. The model estimated net primary production is highest in deep
256 waters near the coast, with production falling to low values at the model boundary (Fig. 3),
257 consistent with the oligotrophic conditions expected for the central Bay (Martin and Shaji,
258 2015). The exception is some areas to the north and west of the domain, where production at
259 the boundary is about half that of the maximum level; the model may miss some production
260 in this part of the domain.
261
262 Projections of change for the 21 st century show a steady rise in sea surface temperature, but a
263 more mixed picture for net primary production (Fig. 3 and Table 1). For the Q0 (standard)
264 climate run, there is a fall in net primary production across most of the northern Bay of
265 Bengal, except for a small area in the main Meghna river mouth. The overall effect for the
266 Bangladesh EEZ is to leave net primary production largely unchanged (<5% change).
267 Primary production in the EEZ rises in both of the warmer runs, Q8 and Q16, and is greatest
268 for Q8 in spite of decreasing river nitrate and phosphate inputs. In the wider bay the general
269 tendency in the longer term is for decreasing production on the Indian coast, but rising
270 production in the Andaman Sea to the east; however, the Indian coast shows rising production
271 at the start of the century for Q8 and Q16. The temperature rise is greatest for the Q16 climate
272 run, which has the highest climate sensitivity: a sea surface temperature rise of nearly 3°C for
273 the Bangladesh area, compared to 2.3°C for Q0.
274 275 In order to investigate the impacts of changes on fish and fisheries potential we ran a size-
276 based model to compute the total potential fish production capacity from 2000 to 2100, in the
277 absence of fishing (Fig. 4). This model projects the transfer of energy from primary to
278 secondary and tertiary producers based on metabolic theory (Blanchard et al. 2012). It has the
279 advantage of having relatively limited parameter demands while providing good estimates of
280 potential catch from the EEZ, which emphasizes the dominant role of body size in accounting
281 for patterns of predatory interaction and production (Blanchard et al. 2012). The results (Fig.
282 4 and Table II) indicate that all three climate runs (all associated with an A1B greenhouse gas
283 emission scenario) project declines in fish productivity. As would be expected, a significant
284 inter-annual and inter-scenario variability is observed. Averaging the results per decade shows
285 that the total fish productivity would decline between 2.6% (Q8 – intermediate temperature
286 rise) and 8.3% (Q0 – standard run) by 2100, depending on the climatic ensemble run
287 considered (Table II). However, the trajectory of this decrease is different for each scenario:
288 steady, significant decrease over time under the Q0 scenario, moderate variations (increase
289 followed by a decrease) until the mid-century under the Q8 scenario, and moderate decrease
290 after the mid-century under the Q16 scenario (Fig. 3). The results for the Bangladesh EEZ
291 domain are consistent with results for the entire BoB shelf region (Table II).
292
293 The weakness of the size-based model is that it does not provide predictions of production
294 and thus potential catches from individual species, and thus does not account for their
295 specific responses to environmental change and fishing. In order to explore the impact of
296 management scenarios we conducted runs using the species-based SS-DBEM model for the
297 two major marine captured species, Hilsa shad and Bombay duck. Given the small difference
298 in overall fish production between the climate ensemble runs (Table II), and to facilitate
299 analysis, we averaged outputs of the three climate ensembles in subsequent species-specific 300 runs and focused on the relative impacts of fisheries management decisions. The results
301 indicate that both Hilsa shad and Bombay duck catches are expected to decline over time
302 regardless of the fisheries management regime, but to different degrees (Fig. 5). For Hilsa the
303 decline stabilises under MSY considerations at 175,000t by 2035, while it virtually collapses
304 around the same period under overfishing (OF) scenarios. A significant inter-annual
305 variability is observed (Fig. 5a). By the 2050s the decline in catches is between 39% (under
306 MSY) and 87% (under OF, Table III). For Bombay duck inter-annual variability is reduced
307 but climate ensemble variability increases the size of the error bars (Fig. 5b). Catch potential
308 declines continuously under all management scenarios but without biological collapse. By the
309 2050s this decline is around 35% for all management scenarios (Table III). It must be noted
310 that the projected catches of both species at the start of the presented time series (2010) differ
311 between management scenarios. This is because fishing mortality rates were applied in the
312 model starting in 1980, to allow populations to stabilise in response to these rates. Thus, by
313 2010 each management scenario is already delivering significant differences in catch
314 potential. Potential catches are on average higher in the more sustainable scenarios (MSY)
315 than in the Business and Usual (BaU) scenarios for both species (91% in Hilsa Shad and 37%
316 in Bombay Duck by the 2050s) (Table III). The less sustainable scenario (Overfishing, OF)
317 projects a decrease of catches of Hilsa shad to <20% of the BaU potential by 2050s (Table
318 III).
319
320 If we compare the outcomes of the current “Business as Usual” (BaU) scenario in the 2000s
321 (2000-2009) period with a more sustainable scenario in coming decades, we expect the
322 management change to result in a very minor decline in potential catches by the 2020s but a
323 still significant (25%) decline by the 2050s (Table IV). Conversely, a future that follows BaU
324 with an overexploitation scenario will bring catches by 2050s almost 95% lower than in the 325 BaU scenario for the start of the 21 st century (Table IV). For Bombay duck, potential catches
326 by 2020s under an MSY scenario will produce over 20% more fish than under BaU at the
327 start of the century, with a decline of less than 20% by the end of the projection period. (Table
328 IV). Conversely, maintaining BaU fishing to the end of century would result in 40% decline
329 in catch potential. This demonstrates that management can mitigate or exacerbate the effects
330 of climate change on ecosystem productivity.
331
332
333 334 Discussion 335
336 The results presented constitute the first effort to project fish and fisheries catch potential in
337 the Bangladesh EEZ over the 21 st century, combining state of the art regional climate models,
338 associated river run-off and nutrient loading volumes, a shelf-specific biochemical model and
339 two conceptually distinct high trophic level model frameworks. The results also assess the
340 subsequent impacts of a number of plausible management scenarios in order to investigate
341 the consequences of global environmental change and management options on the
342 sustainability and food provision of Hilsa shad and Bombay duck.
343
344 Climate change showed a clear impact in the model projections, with sea surface
345 temperatures rising by 2-3° C over the 21 st century and the biggest increases generally in the
346 north of the BoB. The change in net primary production for the whole domain is 3% or less
347 (Table I), in line with the near-static values found by Bopp et al. (2013) in a global study and
348 Blanchard et al (2012) for the Bay of Bengal. This overall small change hides sub-regional
349 differences, with the east and west of the BoB showing different, and scenario-dependent
350 patterns (Fig. 3); further modelling studies, with more up to date climate scenarios and a
351 model domain that includes more of the Bay, would be needed to fully characterize this
352 variation. Net primary production is positively correlated to temperature in the north-western
353 part of the model domain, negatively correlated in the south. Production is weakly correlated
354 to nitrate and phosphate values in the coastal and delta areas but strongly correlated for the
355 deepest areas, suggesting that production is nutrient limited in the central Bay. Production in
356 much of the Andaman Sea is correlated to nutrient values but not to temperature, suggesting
357 that this shallow area is also nutrient-limited, but temperature seems to be a stronger factor
358 for the GBM delta.
359 360 Projected change in net primary production for the Bangladesh EEZ is in the range 0-5% and,
361 as for the wider bay, there are scenario-dependent areas of positive and negative change. Two
362 of the three climate runs tested showed a rising trend in primary production, with the standard
363 climate run the exception. The highest primary production was associated with reduced river
364 flows and nutrient loadings, suggesting that the increase is more affected by changing
365 temperature and sea dynamics than by river inputs. This is supported by analysis of
366 correlations between the modelled nutrient levels and net primary production, which show a
367 relatively weak link in the area of the EEZ.
368
369 In spite of the slight increase in primary production, the projections show a decrease in the
370 potential fish production in the Bangladesh EEZ, of 0-10%. This is because at a higher
371 temperature the steepness of the primary production and fish abundance relationship will
372 increase. This means that there would be higher proportion of fish biomass of smaller size,
373 but lower total fish biomass. As an example using a simple size-spectrum approach that uses
374 only temperature and primary production (Jennings et al., 2008), at the same primary
375 production level, an increase of 2 degrees in temperature can mean a 20% decrease in total
376 biomass. Our model is more complex as it uses different phytoplankton and zooplankton
377 functional groups and additional prey-predator relationships (Blanchard et al., 2012).
378 Blanchard et al. (2012) project declines of 30-60% in tropical shelf and upwelling areas and
379 an increase of pelagic predators by ca. 30-90% in high latitude shelf areas. This temperature
380 driven effect has been also suggested for the Mediterranean Sea (Lejeusne et al., 2010) where
381 temperatures have steadily increased. Hilsa migration is synchronous with the increase of the
382 average water flow and increase in average temperature of river water (Haroon, 1998;
383 Bhaumic et al., 2011). The preferred (optimum) temperature for Hilsa in the breeding ground 384 is 29.3-30.2°C and in the nursing ground is 29.8-30.8°C. The surface temperature is the
385 principal factor governing the distribution of Bombay duck in Bay of Bengal (Bapat 1970).
386
387 Official statistics reveal that marine and inland fish catches in Bangladesh have doubled since
388 1995 reaching 1.6 million tonnes per year, of which Hilsa has contributed ca. 350,000 tonnes
389 (Fig. 6). Over the same period, the number of marine fishing boats and gears has increased
390 about 4 times resulting in tremendous pressure on Hilsa populations. In addition, the intensity
391 of marine catches has increased due to the introduction of nylon twine and mechanized boats
392 (Rahman et al., 2012). Given that our results indicate no evidence of increased productivity,
393 we conclude that such increase in catches is due to the increase in demand for fish on the
394 Bangladeshi and global markets. Although, the population in the South-West coastal zone has
395 not increased significantly between 2001 and 2011 (from 13.9 million to 14.1 million, Szabo
396 2015), because of rural to urban migration, the total population of Bangladesh has steadily
397 increased from 135 to 153 million in the same period (2001-2011) and from 120 million to
398 158 million people during the 1995-2014 period (World Bank 2015). Fish takes up the
399 majority of the animal protein intake in Bangladesh and its consumption has increased from
400 42.1 (2005) to 49.5 (2010) grams per capita per day (BBS 2011). This increase in fish intake
401 was greater in urban areas (49.6 to 59.9 g per capita per day). Thus, the increasing domestic
402 population coupled with changes in consumption levels and urbanisation might be
403 responsible for the increasing commercial and subsistence fishing pressure.
404 The difference in trends between fish landings in the recent past (2000-2015, Fig. 6) and
405 model projections over the same period (Fig. 5) deserves clarification. As mentioned above,
406 historical landings reflect more than ecosystem and fish productivity, especially when
407 management is not sophisticated enough to respond to interannual productivity cycles. There
408 is no detailed spatial and temporal information available in Bangladesh to estimate the annual 409 level of fishing pressure that resulted in the reported recent fish landings. In our simulations,
410 however, we fix fishing mortality over the time of the simulation according to our defined
411 management scenarios (Fig. 5). Thus, the trend and magnitude of the simulated catches
412 within a single simulation reflects ecosystem productivity changes, while differences between
413 simulations show the impact of a fixed fishing pressure on top of the effects of environmental
414 variability. These trends are therefore not expected to match past landings, although their
415 average is consistent with the yield obtained in the Business as Usual scenario.
416
417 Larger decreases are projected for the two main species compared to the total fisheries
418 productivity change from the size-spectrum model, even under sustainable management
419 practices. This apparent discrepancy can be caused by the potential replacement of species
420 such as Hilsa shad and Bombay duck by alternative species, thus resulting in more stable total
421 potential catches at ecosystem level. In the case of Hilsa Shad there are two fish species that
422 have the potential to replace Hilsa Shad: Chacunda Gizzard Shad ( Anodontostoma chacunda )
423 and Toli Shad ( Tenualosa ilisha ). Due to their morphological similarity they are often
424 confused with Hilsa on the markets, particularly if they are young, and are often sold as
425 Hilsa, even though the real value of an adult is three times lower. It is known that the
426 complexity of ecological interactions in the marine food web makes it difficult to extrapolate
427 studies on individual species to community or ecosystem level (Walther et al. 2002). Thus, it
428 is quite possible to observe differential impacts at the community level compared to species
429 patterns. The second cause for this discrepancy is structural: overall fisheries potential is
430 computed using a size-based model, while individual species responses rely on a species-
431 based model. However, both models are forced with the same environmental data, and
432 previous comparisons suggest that both model frameworks tend to have comparable outputs
433 at the right scale (Barange et al. 2014). 434 Model runs that combine environmental change and management considerations demonstrate
435 that the management options to be followed in coming decades are crucial for the
436 sustainability of fisheries and their role as a nutritional and economic resource for the
437 country. Hilsa shad is the largest fishery by volume, and a species with significant economic
438 and cultural value. We showed that the implementation of sustainable management practices
439 would stabilise the marine catch to approx.170,000 tonnes by the 2050s, at 70% of current
440 catches, is likely to reduce Hilsa catches by up to 90% by 2050. A decrease in Hilsa shad
441 catch of this order would have important consequences. While some fish sellers will be able
442 to weather the storm by focusing on other alternative species, it is expected that the impact
443 will be more severe for fishers that have adapted their practices to the dynamics of this
444 species and its consumers’ demand. As many of these fishers depend heavily on Hilsa any
445 decline in Hilsa population will result in poorer living conditions for the people engaged in
446 fishing. The cultural value (Urquhart et al., 2011) and unique taste of Hilsa, which drive its
447 higher value and demand, make it unlikely that the loss of catches for this species will be
448 compensated by fishing other species. This might cause a shift in the workforce from fishing
449 to other livelihood options (Hossain et al., 2013; Nicholls et al., 2013). The loss of fishing
450 livelihood may thus lead to migration to already over-crowded large cities in search of
451 alternative livelihoods.
452
453 Bombay duck productivity is around half that of Hilsa shad, but the catches and biomass are
454 more stable. Our projections indicate that catches of Bombay duck may not collapse as a
455 result of unsustainable fishing practices and climate effects by 2060, contrary to projections
456 for Hilsa. Sustainable practices would maintain current catches to a large extent. The reason
457 for the different response compared to Hilsa relies on the fact that Hilsa shad has slightly
458 higher estimated intrinsic population growth rates and adult movement rate (FishBase 459 database; http://www.fishbase.org/ ) compared to Bombay duck, which makes it more
460 abundant and widely distributed, allowing it to track environmental changes more closely but
461 this also means that it is more sensitive to environmental and climate conditions. In addition,
462 Bombay duck has a higher range of feeding options, from zooplankton, fish larvae and
463 shrimp to other fish species of smaller size (Zhang and Jin, 2014); whereas Hilsa shad is
464 mostly herbivorous though it does eat crustaceans (Dutta et al., 2014). Thus Bombay duck
465 seems to be more resilient to changes in the marine food chain.
466
467 Despite the complexity of the modelling framework, there are still processes that may be
468 under-represented, such as top-down effects (Kearney et al., 2012). Top-down pressures in
469 terms of predator mortality is represented in the models as part of the natural mortality. There
470 is some data about natural mortality of Hilsa, which had a mean value of 1.24 ± 0.08 in the
471 period between 1992 and 2006 (Ahmed et al., 2008; BOBLME, 2010), however this is
472 insufficient to say something quantifiable about top-down effects. Moreover, Hilsa is a highly
473 fecund fish enabling it to compensate for any great loss of progeny which may occur due to
474 predation and unfavorable hydrological conditions (Raja, 1985; de Graaf, 2003; Rahman and
475 Cowx, 2006). There are several piscivorous species that have the potential to predate Hilsa
476 and Bombay Duck (Bahadur, 2010; Murugan et al., 2012) such as sharks, Indian mottled eel
477 (Anguilla bengalensis ), narrow-barred Spanish mackerel ( Scomberomorus commerson ) tunas
478 (Kasuwonus pelamis , Thunnus albacares ,Thunnus obesus ), dolphins (Platanista gangetica
479 gangetica ), seabirds and tooth whales (Nelson, 1998). However, top-predator studies (e.g.
480 stomach content) in this area are rare (Sumontha et al., 2008; Thangavelu et al., 2012;
481 Mazumder et al., 2014). In Sumontha et al. (2008), fish represented 38.9% of the stomach
482 content of several tuna species, the rest were cephalopod. Thangavelu et al. (2012) shows
483 predation of Greater lizardfish ( Saurida tumbil ) on Bombay Duck, but the study is too limited 484 to determine if it is just an opportunistic predation. These fish species are all also fishing
485 targets (Bahadur, 2010; DoF, 2013) and therefore susceptible to human pressures similar to
486 their prey.
487
488 We have not considered in the model the impact that changes in the extension and
489 composition of mangroves would have on fish production. Although the importance of
490 mangroves as fish habitats and nursery grounds are recognized (Hoq, 2003; Chowdhury et al.,
491 2010; Hutchison et al. , 2014), the exact impacts of changes are still uncertain and
492 unquantified. Current research efforts aimed at the kind of quantification that would be
493 needed for modelling relate only to indirect factors. For example, Hutchison et al. (2014)
494 review work related to the impacts on primary production (Alonsegi, 2009; Harrison et al.,
495 1997; O’Donohue and Dennison, 1997). Specific experimental and monitoring work would
496 be needed before we can understand the role of mangroves in their ecosystems further than
497 what-if scenarios.
498 The present study is part of a much broader analysis of coastal Bangladesh. The fisheries
499 model and its results will be used in the integrated analytical framework of the Ecosystem
500 Services of Poverty Alleviation In Populous Deltas (ESPA Deltas, www.espadelta.net)
501 project, which will allow the detailed analysis of not only the fishery resources, but also the
502 fishing livelihoods and its interactions with other provisioning ecosystem services such as
503 agriculture. Since fishing is the second most important source of livelihood in coastal
504 Bangladesh, the integrated model will allow the exploration of the future trajectories of the
505 fishery sector through testing plausible scenarios as a tool to aid sustainable resource
506 management and regional development planning.
507 508 509 Acknowledgements
510 This work ‘Assessing health, livelihoods, ecosystem services and poverty alleviation in
511 populous deltas', project number NE-J002755-1, was funded with support from the
512 Ecosystem Services for Poverty Alleviation (ESPA) programme. The ESPA programme is
513 funded by the Department for International Development (DFID), the Economic and Social
514 Research Council (ESRC) and the Natural Environment Research Council (NERC).We
515 express our gratitude to Julia Blanchard for providing the size-spectrum model code. This
516 work made use of the facilities of HECToR, the UK's national high-performance computing
517 service, which is provided by UoE HPCx Ltd at the University of Edinburgh, Cray Inc and
518 NAG Ltd, and funded by the Office of Science and Technology through EPSRC's High End
519 Computing Programme.
520
521 522 References
523 Ahmed, M.S., Sharif, A.S.M., and Latifa, G.A. 2008. Age, growth and mortality of hilsa shad, Tenualosa ilisha
524 in the river Meghna, Bangladesh. Asian Journal of Biological Sciences, 1: 69-76.
525 Allison, E.H., Perry, A.L., Badjeck, M.C., Adger, N.W., Brown, K., Conway, D., Halls, A.S., Pilling, G.M.,
526 Reynolds, J.D., Andrew, N.L., and Dulvy, N. K. 2009.Vulnerability of national economies to the impacts
527 of climate change on fisheries. Fish and Fisheries , 10: 173-196.
528 Amin, S.M.N., Rahman, M.A., Haldar, G.C., Mazid, M.A., and Milton, D.A. 2008. Catch per unit effort,
529 exploitation level and production of hilsa shad in Bangladesh waters. Asian Fisheries Science, 21: 175-
530 187.
531 Bahadur, H.I. 2010. Trades of shark products in Bangladesh. pp. 43-50. In: Hoq et al. (eds.) 2011. Shark
532 fisheries in the Bay of Bengal, Bangladesh: Status and potentialities. Support to Sustainable Management
533 of the BOBLME Project, Bangladesh Fisheries Research Institute, Bangladesh. 76 p.
534 Bapat, S. V. 1970. The Bombay duck, Harpodon nehereus (Ham.). CMFRI Bulletin, 21: 1-66.
535 Barange, M., Merino, G., Blanchard, J. L., Scholtens, J., Harle, J., Allison, E. H., and Jennings, S. 2014. Impacts
536 of climate change on marine ecosystem production in societies dependent on fisheries. Nature Climate
537 Change, 4: 211-216.
538 Bangladesh Bureau of Statistics (BBS) 2011. Report of the Household Income & Expenditure Survey 2010.
539 Statistical Division, Ministry of Planning.
540 Blanchard, J.L., Jennings, S., Holmes, R., Harle, J., Merino, G., Allen, J.I., Holt, J., Dulvy, N.K., and Barange,
541 M. 2012. Potential consequences of climate change for primary production and fish production in large
542 marine ecosystems. Philosophical Transactions of the Royal Society B, 367: 2979-2989.
543 Blackford, J. C., Allen, J. I., and Gilbert, F. J. 2004. Ecosystem dynamics at six contrasting sites: a generic
544 modelling study Journal of Marine Systems, 52: 191-215.
545 Bay of Bengal Large Marine Ecosystem (BOBLME) project 2010. Status of Hilsa (Tenualosa ilisha)
546 management in the Bay of Bengal. BOBLME-2010-Ecology-01, p. 67.
547 Bopp, L., Resplandy, L., Orr, J.C., Doney, S.C., Dunne, J.P., Gehlen, M., Halloran, P., Heinze, C., Ilyina, T.,
548 Séférian, R., Tjiputra, J., and Vichi, M. 2013. Multiple stressors of ocean ecosystems in the 21st
549 century: projections with CMIP5 models. Biogeosciences 10: 6225-6245. 550 Boyer, T.P., Antonov, J.I., Baranova, O.K., Coleman, C., Garcia, H.E., Grodsky, A., Johnson, D.R., Locarnini,
551 R.A., Mishonov, A.V., O’Brien, T.D., Paver, C.R., Reagan, J.R., Seidov, D., Smolyar, I.V., and Zweng,
552 M.M. 2013. World Ocean Database 2013, NOAA Atlas NESDIS 72. National Oceanographic Data
553 Center Ocean Climate Laboratory, Silver Spring, MD.
554 Caesar, J., Janes, T., Lindsay, A., and Bhaskaran, B. 2015.Temperature and precipitation projections over
555 Bangladesh and the upstream Ganges, Brahmaputra and Meghna systems. Environmental Science:
556 Processes and Impacts, 17: 1047-1056.
557 Cheung, W.W.L., Lam, V.W.Y., and Pauly, D. 2008. Modelling Present and Climate-Shifted Distribution of
558 Marine Fishes and Invertebrates. Fisheries Centre Research Report 16 (3), University of British
559 Columbia, Vancouver.
560 Cheung W. W. L., Close C., Kearney K., Lam, V. W. Y., Sarmiento J., Watson R., and Pauly D. 2009. Projections
561 of global marine biodiversity impacts under climate change scenarios. Fish and Fisheries, 10: 235-251.
562 Cheung, W. W. L., Lam, V. W. Y., Sarmiento, J. L., Kearney, K., Watson, R. E. G., Zeller, D., and Pauly, D.
563 2010. Large ‐scale redistribution of maximum fisheries catch potential in the global ocean under climate
564 change. Global Change Biology, 16: 24-35.
565 Cheung, W. W. L., Dunne, J., Sarmiento, J. L., and Pauly, D. 2011. Integrating ecophysiology and plankton
566 dynamics into projected maximum fisheries catch potential under climate change in the Northeast
567 Atlantic. ICES Journal of Marine Science, 68: 1008-1018.
568 Close, C., Cheung, W. W. L. , Hodgson, S., Lam,V. W. Y., Watson, R., and Pauly, D. 2006. Distribution ranges of
569 commercial fishes and invertebrates. In: Palomares MLD, Stergiou KI, Pauly D (eds) Fishes in databases
570 and ecosystems. Fisheries Centre Research Reports 14 (4), Fisheries Centre, University of British
571 Columbia, Vancouver, p 27–37.
572 de Graaf, G.J, 2003, Water management and the drift of larval fish in the floodplains of Bangladesh. Practical
573 experiences of the Compartmentalization Pilot Project with "fish friendly" regulators. Technical report
574 Nefisco foundation. Amsterdam, The Netherlands.
575 Dee, D.P., Uppala, S.M., Simmons, A.J., Berrisford, P., Poli, P., Kobayashi, S., Andrae, U., Balmaseda, M.A.,
576 Balsamo, G., Bauer, P., Bechtold, P., Beljaars, A.C.M., van de Berg, L., Bidlot, J., Bormann, N., Delsol,
577 C., Dragani, R., Fuentes, M., Geer, A.J., Haimberger, L., Healy, S.B., Hersbach, H., Hólm, E.V., Isaksen,
578 L., Kållberg, P., Köhler, M., Matricardi, M., McNally, A.P., Monge-Sanz, B.M., Morcrette, J.-J., Park, B.- 579 K., Peubey, C., de Rosnay, P., Tavolato, C., Thépaut, J.-N., and Vitart, F. 2011. The ERA-Interim
580 reanalysis: configuration and performance of the data assimilation system. Quarterly Journal of the Royal
581 Meteorological Society, 137: 553-597.
582 Delgado, C.L., Wada, N., Rosegrant, M.W., Meijer, S., Ahmed, M., 2003. Fish to 2020: Supply and Demand in
583 Changing Global Markets. Washington, DC (US) and Penang (Malaysia).
584 DoF, 2013.National Fish Week 2013 Compendium (in Bengali).Department of Fisheries, Ministry of Fisheries
585 & Livestock, Bangladesh, 144 p.
586 Dutta, S., Maity, S., Chanda, A., and Hazra, S. 2012. Population structure, mortality rate and exploitation rate of
587 hilsa shad ( Tenualosa ilisha ) in West Bengal coast of northern Bay of Bengal, India. World Journal of
588 Fish and Marine Sciences, 4: 54-59.
589 Dutta, S., Maity, S., Bhattacharyya, S. B., Sundaray, J. K., and Hazra, S. 2014. Diet Composition and Intensity
590 of Feeding of Tenualosa ilisha (Hamilton, 1822) Occurring in the Northern Bay of Bengal, India. In
591 Proceedings of the Zoological Society (Vol. 67, No. 1, pp. 33-37). Springer India.
592 FAO, 2007.The State of World Fisheries and Aquaculture - 2006. Food and Agriculture organization. Rome,
593 Italy.
594 Fernandes, J. A., Cheung, W. W., Jennings, S., Butenschön, M., Mora, L., Frölicher, T. L, Barange M., and
595 Grant, A. 2013. Modelling the effects of climate change on the distribution and production of marine
596 fishes: accounting for trophic interactions in a dynamic bioclimate envelope model. Global change
597 biology, 19: 2596-2607.
598 Ferry, N., Parent, L., Garric, G., Drevillon, M., Desportes, C., Bricaud, C., and Hernandez, F., 2012. Scientific
599 Validation Report (ScVR) for Reprocessed Analysis and Reanalysis. MyOcean project report, MYO-
600 WP04-ScCV-rea-MERCATOR-V1.0 (No. WP 04 – GLO – MERCATOR – TOULOUSE - FR). Toulouse,
601 France.
602 Froese, R., Zeller, D., Kleisner, K., and Pauly, D. 2012. What catch data can tell us about the status of global
603 fisheries. Marine biology, 159: 1283–1292.
604 Garcia, H.E., Locarnini, R.A., Boyer, T.P., Antonov, J.I., Zweng, M.M., Baranova, O.K., and Johnson, D.R.,
605 2010. World Ocean Atlas 2009, Volume 4: Nutrients (phosphate, nitrate, silicate). S. Levitus, Ed. (No.
606 NOAA Atlas NESDIS 71).U.S. Government Printing Office, Washington, D.C. 607 Haldar, G.C., and Amin, S. M. N. 2005.Population dynamics of male and female hilsa, Tenualosa ilisha of
608 Bangladesh. Pakistan Journal of Biological Science, 8: 307-313.
609 Haroon, A.K.Y. 1998. Hilsa shad: Fish for the teeming millions, new management alternatives needed for the
610 Hilsa young. Shad Journal 3, 7-10.
611 Hilborn R., and Walters, C.J. 1992. Quantitative fisheries stock assessment: choice, dynamics and uncertainty.
612 Chapman & Hall, New York.
613 Hollmann, R., Merchant, C.J., Saunders, R., Downy, C., Buchwitz, M., Cazenave, A., Chuvieco, E., Defourny,
614 P., de Leeuw, G., Forsberg, R., Holzer-Popp, T., Paul, F., Sandven, S., Sathyendranath, S., van
615 Roozendael, M., and Wagner, W. 2013. The ESA Climate Change Initiative: Satellite Data Records for
616 Essential Climate Variables. Bulletin of the American Meteorological Society, 94: 1541-1552.
617 Holt, J. T., and James, I. D. 2001. An s coordinate density evolving model of the northwest European continental
618 shelf: 1 model description and density structure Journal of Geophysical Research: Oceans (1978–2012),
619 106: 14015-14034.
620 Holt, J., Harle, J., Proctor, R., Michel, S., Ashworth, M., Batstone, C., Allen, I., Holmes, R., Smyth, T., Haines,
621 K., Bretherton, D., and Smith, G. 2009. Modelling the global coastal ocean. Philosophical Transactions
622 of the Royal Society A, 367: 939-951.
623 Hossain, M. S., Hein, L., Rip, F. I., and Dearing, J. A. 2013. Integrating ecosystem services and climate change
624 responses in coastal wetlands development plans for Bangladesh. Mitigation and Adaptation Strategies
625 for Global Change, 1-21.
626 Hussain, M.G., and Hoq, M.E. (eds.). 2010. Sustainable Management of Fisheries Resources of the Bay of
627 Bengal- Compilation of national and regional workshop reports. Support to Sustainable Management of
628 the BOBLME Project, Bangladesh Fisheries Research Institute. SBOBLMEP Pub./Rep. 2. 122 p.
629 Hutchison J, Spalding M, and zu Ermgassen P. 2014. The role of mangroves in fisheries enhancement. The
630 Nature Conservancy and Wetlands International, UK.
631 Intergovernmental Panel on Climate Change (IPCC) 2007. Climate Change 2007: The Physical Science Basis.
632 Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on
633 Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA. 634 Jennings, S., Mélin, F., Blanchard, J. L., Forster, R. M., Dulvy, N. K., and Wilson, R. W. 2008. Global-scale
635 predictions of community and ecosystem properties from simple ecological theory. Proceedings of the
636 Royal Society of London B: Biological Sciences, 275: 1375-1383.
637 Kearney, K. A., Stock, C., Aydin, K., and Sarmiento, J. L. 2012. Coupling planktonic ecosystem and fisheries
638 food web models for a pelagic ecosystem: Description and validation for the subarctic Pacific. Ecological
639 Modelling, 237: 43-62.
640 Khan, M. S. U. 2007. Optimal stock, harvest and effort level of Bangladesh trawl shrimp fishery – a nonlinear
641 dynamic approach. Journal of Agriculture & Rural Development, 5: 143-149.
642 Lejeusne, C., Chevaldonné, P., Pergent-Martini, C., Boudouresque, C. F., and Perez, T. 2010. Climate change
643 effects on a miniature ocean: the highly diverse, highly impacted Mediterranean Sea. Trends in ecology
644 & evolution: 25, 250-260
645 Martin, M.V., and Shaji, C. 2015. On the eastward shift of winter surface chlorophyll-a bloom peak in the Bay
646 of Bengal. Journal of Geophysical Research: Oceans, 120: 2193-2211.
647 Mazumder, M. K., Boroa, F., Barbhuiya, B. and Singha, U. 2014. A study of the winter congregation sites of
648 the Gangetic River Dolphin in southern Assam, India, with reference to conservation. Global Ecology
649 and Conservation, 2: 359-366
650 McAllister, M. K., Pikitch, E. K., and Babcock, E. A. 2001.Using demographic methods to construct Bayesian
651 priors for the intrinsic rate of increase in the Schaefer model and implications for stock rebuilding.
652 Canadian Journal of Fisheries and Aquatic Sciences, 58: 1871-1890.
653 Murugan, S., Joseph A. P. and Khan S. A. 2012. Ecological niche of Mugil Cephalus – an ECOPATH with
654 ECOSIM approach in Vellar Estuary (South east coast of India). International Journal of Pharma and Bio
655 Sciences, 3: S662-S676.
656 Newton, K., Cote, I. M., Pilling, G. M., Jennings, S., and Dulvey, N. K. 2007.Current and Future Sustainability
657 ofIsland Coral Reef Fisheries. Current Biology, 17: 655-658.
658 Nicholls, R.J., Hutton, C.W., Lazar, A.N., Rahman M., Salehin, M., Ghosh, T. 2013. Understanding Climate
659 Change and Livelihoods in Coastal Bangaldesh. Hydrolink, 2, pp. 40.
660 Nicholls, R.J., Whitehead, P., Wolf, J., Rahman, M., Salehin, M., 2015. The Ganges–Brahmaputra–Meghna
661 delta system: biophysical models to support analysis of ecosystem services and poverty alleviation.
662 Environmental Science: Processes and Impacts, 17: 1016-1017. 663 Pauly, D. 1980. On the interrelationships between natural mortality, growth parameters and mean environmental
664 temperature in 175 fish stocks. ICES Journal of Marine Science, 39: 175-192.
665 Pauly, D., Hilborn, R., and Branch, T. A. 2013. Fisheries: does catch reflect abundance? Nature, 494: 303-306.
666 Rahman, M.A., Alam, M.A. Hasan, S.J. and Zaher, M. 2012. Hilsa (Tenualosa ilisha) Fishery Management in
667 Bangladesh. In: Hilsa: Status of Fishery and Potential for Aquaculture (ed. Anon), Proceedings of the
668 Regional Workshop held in Dhaka, 16–17 September 2012. The WorldFish, Bangladesh and South Asia
669 Office, Dhaka, Bangladesh, 40-61 pp.
670 Rahman, M.J., and Cowx, I.G. 2006. Lunar periodicity in growth increment formation in otoliths of Hilsa shad
671 (Tenualosa ilisha, Clupeidae) in Bangladesh waters. Fisheries Research, 81: 342-344.
672 Raja, B.T.A. (1985) A review of the biology and fisheries of Hilsa ilisha in the Upper Bay of Bengal.Bay of
673 Bengal Program.BOBP/WP/37.Marine Fishery Resource Management in the Bay of Bengal. p. 66.
674 Rao, K. 2013. Bombay duck: iconic fish fast disappearing from city's coastal waters. The Guardian Friday 22
675 March 2013.
676 Szabo, S., Begum, D., Ahmad, S., Matthews, Z., and Streatfield, P.K. 2015. Scenarios of population change in
677 the coastal Ganges Brahmaputra delta (2011 - 2051). ESRC Centre for Population Change, Working
678 Paper 61, March 2015, ISSN 2042-4116.
679 Schaefer, M. 1954. Some aspects of the dynamics of populations important to the management of the
680 commercial marine fisheries. Bull I-ATTC, 1: 27-56.
681 Shamsuddoha, M. and Chowdhury, M. R. K. 2007.Climate Change Impact and Disaster Vulnerabilities in the
682 Coastal Areas of Bangladesh. COAST Trust, House 9/4, Road 2, Shaymoli, Dhaka, Bangladesh
683 Sparre, P., and Venema, S. C. 1992.Introduction to tropical fish stock assessment. Part 1.Manual.FAO Fisheries
684 Technical Paper 306, FAO, Rome.
685 Sumontha, M., Nootmorn, P., Keereerut, P., Jayasinghe, R.P.P.K., Jagannath, N. and Sinha M.K. 2008. Stomach
686 Content of the Large Pelagic Fishes in the Bay of Bengal. The Ecosystem-Based Fishery Management in
687 the Bay of Bengal: 206-220 pp. IOTC-2008-WPEB-11.
688 Thangavelu, R., Anbarasu, M., Zala, M. S., Mohamed Koya, K., Sreenath, K. R., Mojjada, S.K. and Shiju P.
689 2012. Food and feeding habits of commercially important demersal finfishes off Veraval coast. Indian
690 Journal of Fisheries, 59: 77-87. 691 Toufique, K. A., and Belton, B. 2014. Is aquaculture pro-poor? Empirical evidence of impacts on fish
692 consumption in Bangladesh. World Development, 64: 609-620.
693 Urquhart, J., T. Acott, M. Reed, and P. Courtney. 2011. Setting an agenda for social science research in fisheries
694 policy in Northern Europe. Fisheries Research, 108: 240-247.
695 Walther, G. R., Post, E., Convey, P., Menzel, A., Parmesan, C., Beebee, T. J., Fromentin, J., Hoegh-Guldberg,
696 O., and Bairlein, F. 2002. Ecological responses to recent climate change. Nature, 416: 389-395.
697 Whitehead, P.G., Barbour, E., Futter, M.N., Sarkar, S., Rodda, H., Caesar, J., Butterfield, D., Jin, L., Sinha, R.,
698 Nicholls, R., Salehin, M., 2015a. Impacts of climate change and socio-economic scenarios on flow and
699 water quality of the Ganges, Brahmaputra and Meghna (GBM) river systems: low flow and flood
700 statistics. Environmental Science: Processes and Impacts, 17: 1057-1069.
701 Whitehead, P.G., Sarkar, S., Jin, L., Futter, M.N., Caesar, J., Barbour, E., Butterfield, D., Sinha, R., Nicholls, R.,
702 Hutton, C., Leckie, H.D., 2015b. Dynamic modeling of the Ganga river system: impacts of future climate
703 and socio-economic change on flows and nitrogen fluxes in India and Bangladesh Environmental
704 Science: Processes and Impacts, 17: 1082-1097.
705 World Bank 2015. Total population (in number of people): Bangladesh. Website accessed: 11/08/2015.
706 http://data.worldbank.org/indicator/SP.POP.TOTL/countries/BD?display=graph
707 Zhang, B., and Jin, X. 2014.Feeding habits and ontogenetic diet shifts of Bombay duck, Harpadonnehereus .
708 Chinese Journal of Oceanology and Limnology, 32: 542-548
709 710 Tables 711 Table I
Area Climate Present Mid- End- scenario day century century SST (°C) Bangladesh Q0 28.3 29.5 30.6 precipitation (mm day -1) EEZ 8.6 9.1 9.3 net PP (g C m -2 day -1) 1.65 1.67 1.66 river flow (×10 10 m3) 98 103 104 river nitrate (tonne) 345 321 370 river phosphate (tonne) 50 50 56 SST (°C) BoB Q0 29.5 30.8 31.7 precipitation (mm day -1) 6.9 7.2 7.5 net PP (mg C m -2 day -1) 1.26 1.24 1.24 SST (°C) Bangladesh Q8 28.4 29.9 31.0 precipitation (mm day -1) EEZ 7.2 6.8 7.1 net PP (g C m -2 day -1) 1.68 1.77 1.78 river flow ×10 10 m3) 108 97 109 river nitrate (tonne) 370 319 334 river phosphate (tonne) 56 49 53 SST (°C) BoB Q8 29.5 30.8 31.9 precipitation (mm day -1) 3.2 6.0 5.9 net PP (mg C m -2 day -1) 1.29 1.33 1.29 SST (°C) Bangladesh Q16 29.2 30.8 32.1 precipitation (mm day -1) EEZ 5.8 5.4 5.8 net PP (mg C m -2 day -1) 1.79 1.84 1.83 river flow ×10 10 m3) 94 97 108 river nitrate (tonne) 322 268 267 river phosphate (tonne) 48 52 55 SST (°C) BoB Q16 30.3 31.7 32.9 precipitation (mm day -1) 5.0 4.9 5.1 net PP (mg C m -2 day -1) 1.35 1.33 1.37 712
713 Table I. Illustration of the differences between the three climatic and hydrological scenarios 714 used to drive the oceanic physical-biogeochemical model of the Bay of Bengal. Sea surface 715 temperature (SST), precipitation and net primary production (net PP) data are given for the 716 Bangladesh EEZ region, north of 18 oN, and for the whole BoB (in shaded rows). Present-day, 717 mid- and end-century values are means for 1991-2010, 2041-2060 and 2081-2099 718 respectively. River flow and nutrient loadings are for the Ganges-Brahmaputra-Meghna 719 (GBM) system only. Note that the river flow depends on precipitation in the Ganges and 720 Brahmaputra basins, to the north and west of the model domain, and on patterns of river 721 water extraction, rather than precipitation in the Bay of Bengal.
722
723 724 Table II
Area Climate Fish Class 2020s-2000s 2050s-2000s 2090s-2000s Scenario ∆ change (%) ∆ change (%) ∆ change(%) Bangladesh Q0 Total (<120 cm) -0.6 -4.9 -8.3 EEZ <30cm -0.7 -5.5 -9.1 30-90cm -1.0 -5.3 -9.0 BoB Q0 Total (<120 cm) -3.0 -7.3 -7.7 <30cm -3.2 -8.0 -8.8 30-90cm -2.9 -7.2 -7.7 Bangladesh Q8 Total (<120 cm) -0.7 -2.6 -2.6 EEZ <30cm -0.8 -3.1 -3.4 30-90cm -1.1 -3.3 -3.4 BoB Q8 Total (<120 cm) -0.3 -1.8 -5.3 <30cm -0.8 -2.4 -7.2 30-90cm -0.8 -2.4 -5.6 Bangladesh Q16 Total (<120 cm) -1.8 -1.3 -5.6 EEZ <30cm -2.4 -1.4 -6.5 30-90cm -2.1 -1.6 -6.4 Q16 Total (<120 cm) -1.9 -4.3 -3.5 BoB <30cm -4.4 -5.0 -2.2 30-90cm -1.8 -4.2 -3.7 725
726 Table II. Change in production potential in the 2020s, 2050s and 2090s decades (10-year
727 averages), compared to the 2000s decade, by climate ensemble run. Potential production
728 refers to total fish, and fish of <30cm and between 30-90cm, using a size-spectrum model. 729 Table III
730
2020s-2000s 2050s-2000s 2000s 2020s 2050s
∆ catch (%) ∆ catch (%) Average Average Average
catch (‘000 catch (‘000 catch (‘000
tonnes/yr) tonnes/yr) tonnes)
MSY -20.6 ± 14.5 -39.0 ± 20.1 283.0 ± 221.7 ± 168.5 ± 35.3
30.0 14.9
Hilsa Shad BaU -30.7 ± 12.7 -42.1 ± 7.6 227.9 ± 156.7 ± 101.3 ± 11.9
14.0 19.7
OF -29.3 ± 8.8 -87.2 ± 11.1 165.6 ± 3.0 116.9 ± 15.7 ± 13.6
13.0
MSY -7.4 ± 4 -37.2 ± 4.0 124.2 ± 114.7 ± 77.8 ± 11.4
15.6 10.2
Bombay Duck BaU -13.9 ± 1.9 -32.4 ± 7.0 98.2 ± 23.7 84.4 ± 19.5 57.3 ± 15.4
OF -19.2 ± 1.7 -33.5 ± 7.3 80.3 ± 20.7 65.0 ± 17.3 43.5 ± 13.6
731
732
733 Table III. Change in catch potential for Hilsa shad and Bombay duck in the 2020s and 2050s,
734 referenced to the 2000s decade (in %), and average catch in each of the three decades (t/yr),
735 according to three fisheries management scenarios. The reported values are the mean and
736 standard deviation across Q climate scenarios.
737
738 739 Table IV
740
2020s ∆ catch (%) 2050s ∆ catch (%)
Present BaU to MSY -2.5 ± 7.6 -25.9 ± 14.0
Hilsa Shad Present BaU to BaU -31.1 ± 8.3 -61.1 ± 11.0
Present BaU to OF -48.6 ± 5.6 -93.1 ± 5.2
Present BaU to MSY 20.9 ± 24.1 -18.0 ± 18.4
Bombay Duck Present BaU to BaU -11 ± 24.4 -39.6 ± 18.1
Present BaU to OF -31.5 ± 20.4 -54.7 ± 15.1
741
742 Table IV. Change in catch potential for Hilsa shad and Bombay duck in the 2020s and 2050s,
743 according to three fisheries management scenarios, referenced to the period 2000-2010 under
744 the Business as Usual (BaU) scenario (in %). The reported values are the mean and standard
745 deviation across the three climate ensemble runs.
746
747 748 Figures 749
750 Figure 1. Flowchart and structure of the models used to estimate the impacts and consequences of
751 climate change and socio-economic scenarios on Bangladesh fisheries. The shaded boxes represent
752 the models directly used in this paper.
753
754 Figure 2. Map of the Bay of Bengal showing the POLCOMS-ERSEM model domain
755 (shaded), the position of river mouths in the model and the Bangladesh Exclusive Economic
756 Zone. Diamonds show the position of rivers whose flow and nutrient loadings were estimated
757 from the INCA model; circles indicate river flows and nutrients estimated using global
758 databases (see text for details). Depth contours are shown for 100 and 1000 metres.
759
760 Figure 3. Changes in sea surface temperature (SST) and net primary production (net PP)
761 projected by the Bay of Bengal model for the 2020s, 2050s and 2090s, compared to a
762 baseline of values for the 2000s. The rows show results for the three climate ensemble runs
763 Q0 (top row), Q8 (middle) and Q6 (bottom row). Changes in river flow volume, nitrate (N)
764 and phosphate (P) loads of the GBM rivers are shown in the left hand panel of each pair,
765 these values also hold for the right hand panel. Note that the SST legend changes for each
766 time slice.
767
768 Figure 4. Time series of changes in projected total fisheries production potential in the
769 Bangladesh EEZ according to the three climate ensemble runs (Q0, Q8 and Q16) and in the
770 absence of fisheries extractions. Values are expressed as a percentage deviation from the 2000
771 production for each ensemble run.
772
773 774 Figure 5. Time series of catch potential projections (t) in the Bangladesh EEZ for Hilsa shad (A) and
775 Bombay duck (B) under different fisheries management scenarios (MSY or Sustainable exploitation,
776 BaU or Business as Usual and OF or Overfishing scenario). Error bars indicate variability between the
777 three climate ensemble runs (Q0, Q8 and Q16) and associated river run-off and nutrient loadings.
778
779 Figure 6. Time series of marine and inland catch (Total and for Hilsa shad) and total
780 aquaculture production in Bangladesh since 1950 (source: FAO Global database).
781
782 783 Figure 1 784
785 786 787 Figure 2 788 789
790 791
792 793 Figure 3. 794
795
796 797 Figure 4. 798
799
800 801 Figure 5.
802
803
804
805 806 807 Figure 6
808
809